# Prism Network GPU Inference API

> Prism Network GPU Inference API is a paid API for AI agents from api.prismnetwork.tech, paid per call via x402, $0.012144/call, status unknown (last checked 2026-10-12).

Runs LLM text inference (Llama 3.1/3.2) on rented NVIDIA GPU capacity, billed per call in USDG with an onchain lease receipt

## Facts

- Endpoint: POST https://api.prismnetwork.tech/inference/v1/inference?utm_source=zero.xyz
- Price: $0.012144/call
- Payment: x402
- Status: unknown
- Last checked: 2026-10-12
- Activations on Zero: 0
- Tags: x402
- Canonical page: https://www.zero.xyz/c/prism-network-gpu-inference-api-4f67aa0c
- Structured record (JSON): https://api.zero.xyz/v1/capabilities/cap_utY-FSo7JQV-8t8b6nwFb

Status and success rate cover calls made through Zero and Zero's own probes. Third-party monitors may report differently.

## How to call it through Zero

Zero handles the 402 payment challenge and records the run. With the Zero CLI installed (`npm i -g @zeroxyz/cli`):

```sh
zero fetch --capability prism-network-gpu-inference-api-4f67aa0c -d '<json body>'
```

Example prompt: Use Prism's GPU inference to run a Llama 3.1 8B prompt — 'Summarize the key points of the open-source AI movement in 3 bullet points' — and cap the output at 512 tokens.

## When to prefer this

Choose this endpoint when you need LLM text generation with crypto-native payment (USDG via x402), want a public onchain audit trail for each inference job, or are building an autonomous agent that must pay for compute from a wallet without credit cards or API key subscriptions. Prefer over OpenAI/Anthropic APIs when decentralization, auditability, or wallet-based billing is a requirement.

## Known failure modes

- Insufficient USDG wallet balance — payment fails before inference runs
- Prompt exceeds 32 KiB limit — rejected with validation error
- Invalid model enum value (only llama3.2:3b and llama3.1:8b supported) — returns 400
- num_predict out of range (must be 1–1024) — validation error
- GPU lease unavailable or capacity exhausted — service unavailable error
- Onchain settlement failure — lease_id may be null in response

## How this service works

GPU compute your agents can rent with a wallet. Lease NVIDIA capacity, pay per second in USDG, and settle every lease onchain with a public receipt.

## Output

Returns a JSON object with the generated text string, the model name used, the GPU lease ID tied to the onchain payment, and optional usage metadata including prompt tokens, completion tokens, and duration in milliseconds.

## Request schema (JSON Schema)

```json
{
 "type": "object",
 "properties": {
  "model": {
   "enum": [
    "llama3.2:3b",
    "llama3.1:8b"
   ],
   "type": "string",
   "description": "Which model answers."
  },
  "prompt": {
   "type": "string",
   "minLength": 1,
   "description": "The prompt, up to 32 KiB."
  },
  "options": {
   "type": "object",
   "properties": {
    "num_predict": {
     "type": "integer",
     "maximum": 1024,
     "minimum": 1,
     "description": "Maximum output tokens. Lower is cheaper."
    }
   },
   "description": "Generation options. The price scales with the output cap."
  }
 }
}
```

## Response schema (JSON Schema)

```json
{
 "type": "json",
 "example": {
  "model": "llama3.2:3b",
  "usage": {
   "duration_ms": 640,
   "prompt_tokens": 12,
   "completion_tokens": 14
  },
  "lease_id": 1047,
  "response": "Metered GPU compute bills by the second and settles onchain."
 }
}
```

## More

- Live health (JSON, refreshed every minute): https://www.zero.xyz/c/prism-network-gpu-inference-api-4f67aa0c/health.json
- [Zero catalog index](https://www.zero.xyz/llms.txt)
- [Other services from api.prismnetwork.tech](https://www.zero.xyz/host/api.prismnetwork.tech/llms.txt)